Here’s why…
When humans and machines work together, we get better, faster and safer ways of moving the world forward and we have countless examples throughout history to prove it.
The monumental progress in automation has expanded the capabilities of modern machines to unprecedented levels, giving the human world a considerable leg up in advancing productivity, performance and safety across industrial sectors. This is a good thing.
And yet, we still face growing concern over the job-destroying potential of these innovations, a topic that has only gained momentum recently with the impact of Covid-19 accelerating automation projects.
This isn’t the first time advancements in technology have created a challenge between humans and machines. (No, we are not talking about the Terminator, although a rogue Atlas robot can get pretty close!).
This resistance goes back to the early 1800s when cotton and wool workers rebelled against the power looms, seen as a threat to hand weavers. This type of resistance continued with events like the Swing riots and goes on today with protests against automated checkout stations and self-driving trucks.
Were any of these conflicts successful in rolling back the progress? Not even close.
On the contrary, when you look at robotics throughout history, we repeatedly see how humans and machines working together makes change for the better.
Introducing industrial robots – Hello Knucklehead!
More than 65 years ago, in 1956, George Devol an American inventor and physicist, made an early prototype of what he then called a “Programmed Article Transfer device”. Three years and many iterations later, his invention translated into the first industrial-grade robotic arm, named Unimate. General Motors was the first company to deploy this 2700-pound robot on a diecasting assembly line in Trenton, New Jersey. In a very short period, approximately 450 Unimate robotic arms were deployed in diecasting by large and small manufacturers. Machine operators nicknamed the robot as “Knucklehead”. Manufacturers and Unimation were supported by the worker unions because the new equipment wasn’t used to displace workers. New robots were added only in response to attrition, which was high given the harsh working environment of the diecasting shops.
Almost parallelly, KUKA developed its first automatic welding system and a few years later FANUC shipped its first numerical control (NC) machines. Robotics technology in the decades that followed moved from hydraulic systems to all-electric robots primarily serving the automotive manufacturing industry.
The primary theme of these early innovations was their focus on automating tasks that were too dangerous for humans to continue doing, particularly while pushing the boundaries of productivity. The machine operators largely welcomed the knuckleheads and robotic arms in favour of moving to higher skilled jobs within the firms.
These robots can move…on their own
Move over Elmer, Elsie, Shaky and Flakey and make room for CARMEL! (Autonomously, of course)
The first autonomous robots go back to 1948 when William Grey Walter created two ‘tortoises’: Elmer and Elsie. While they never made their way into industrial applications, they were the cornerstones. Twenty years later, in 1968, the first mobile robot, affectionately named ‘Shakey’ was developed at Stanford Research institute (SRI). Shakey, at one point referred to as the “first electronic person”, had a deep impact on robotics and AI research. Shakey’s successor – Flakey was launched in 1985, but got its butt kicked by the University of Michigan hero, CARMEL, at the first AMR competition held in 1992. Since the days of CARMEL, AMRs have come a long way. NASA, an early adopter of these advanced mobile robots, used them primarily for exploration purposes. Several years later, AMRs made their mainstream debut in the industrial verticals with warehousing applications. The logistics industry quickly started to deploy AMRs to move goods from one point to another. Packed with some of the best technology that exists today, AMRs in industrial environments not only improve efficiency and productivity, but also conditions for workers. AMRs take on repetitive and cumbersome jobs, giving workers an opportunity to move to other more skilled tasks requiring human expertise.
And they’re getting stronger
Imagine bench-pressing a Toyota Camry?
Earlier this month, Clearpath, a McRock Capital portfolio company launched the 2nd generation of its OTTO 1500 AMR. In 2015, they were the first-to-market in the heavy-class AMR. When we say heavy class, the OTTO 1500s really mean business. This thing can lift your Camry and do a bench press with it. Today, AMRs have fundamentally changed industrial material handling workflows and opened doors for deeper robot-to-robot interactions.
AMR platforms like OTTO, equipped with arms and manipulators, can take on far more challenging use cases like mobile picking, truck loading and unloading in verticals like logistics, retail and manufacturing. A glimpse of the possibilities were flaunted at the launch of Stretch systems a few weeks ago.
Yikes! Now they can see
Machine vision as a concept dates back as far as the 1930s, when Electronic Sorting Machines offering food, used sorters based on specific filters and photomultiplier detectors. Modern day vision systems started with Cognex, founded 1981, producing its first vision system, DataMan, an industrial optical character recognition (OCR) system capable of reading and verifying the quality of text marked directly on parts and components. In the same year, the industry also witnessed the first production implementation of a machine vision system right here in Canada. A General Motors foundry at the St. Catherines, Ontario, used Consight-I systems to successfully sort up to six different castings at up to 1,400 an hour from a belt conveyor using three industrial robots. Now, that’s impressive!
Today, “3D vision” is perhaps the most important breakthrough giving robots a more complete understanding of their environments. It also lets robots handle unexpected variables better, helps machines collaborate with humans safely and enables robots to perform new tasks on the fly without constant reprogramming.
Humans and machines – the best of both worlds
Should we be worried about these robots that now have arms, vision, mobility and can also use AI to teach itself? Aren’t they getting a little too close to humans? The short answer is, NO.
There is no way to reduce automation to a simple binary between “manual” and “autonomous.” Instead, it is about finding the right balance between aspects that we would find useful to automate and tasks where it remains meaningful for humans to participate. There are many practical benefits of humans participating in automation systems. Let’s call this human/machine interaction, hybrid systems.
These hybrid systems take away the pressure of building perfect algorithms (if any such thing exists). Using human intelligence and interaction, automation systems are exempted from getting everything right while also making it a quick and efficient deployment. Further, hybrid systems can incorporate human preferences, a flexibility that lacks in otherwise autonomous systems. Finally, an important advantage of such systems is transparency. When humans and AI undertake a task alongside one another, it becomes harder and harder for the process to remain hidden. This transparency helps in getting buy-in from the operators and decision makers alike, a very key requirement for a solution to scale in the industrial sectors.
This hybrid philosophy is at the heart of one of McRock’s portfolio companies, Plus One Robotics, where they believe Robots Work, People Rule! as they pioneer human-in-the-loop (HITL) systems. The company gives robots the eye-hand coordination to pick and place objects in the warehouses and distribution centers using 3D and AI-powered perception, so robots get smarter over time. Plus One-enabled robots can occasionally call for help from digitally connected humans called, Crew Chiefs.
Erik Nieves, Plus One founder says “We believe the future of robots isn’t to try to continue to make robots smarter, but to connect robots to smart people”. Well said, Erik!
Robots taking over the world? Yes. Taking over the role of humans in the world? Not today Wall•E.
As for the “robopocalypse”, we’ll leave that for feature films and science fiction novels.
Authors: Udit Bhatnagar and Akira Tanabe